The objectives of the Adverse Drug Event Detection and Intervention project are to reliably measure and reduce the incidence of adverse drug events (ADEs) suffered by patients within three hospitals of the Duke University Health System using a computerized system for ADE detection, reporting, and intervention. A computerized ADE detection system will be established using information system components largely in place today at DUHS, and a methodology will be implemented for the selective alerting of physicians about critical detected events. A computer event engine that receives transaction data from hospital clinical systems such as laboratory, pharmacy, order management will use programmed rules to screen for "trigger" data that alone or in combination suggest the occurrence of an ADE. A subset of rules will be selected for immediate electronic communication to clinicians caring for the involved patient, permitting evaluation and intervention to prevent or ameliorate a possible ADE. In addition, all rules triggered will send a notification to a report generator and be compiled into a daily report for evaluation by pharmacists trained in ADE investigation. Every triggering event will be investigated in detail within 24 hours to determine whether it represents an ADE. The evaluation will include review of the patient's chart and discussion with clinicians involved in the patient's care and if possible, discussion with the patient. If an adverse event is determined to have occurred the investigator will evaluate its causality, severity, and sequelae, and initiate a procedure for follow-up tracking and data gathering. An electronic report will be completed and entered into a database for further analysis. In addition to permitting immediate intervention and mitigation of ADEs, the automated surveillance system will permit the establishment of baseline statistics on the incidence and nature of ADEs across DUHS. This will permit evaluation of the effectiveness of alert-generated interventions, as well as the effectiveness of other interventions currently in implementation to improve medication safety (for example, computerized physician order entry).